Professional Documents
Culture Documents
DISCUSSION
Since the peptides recognized by TEIPP-specific CTL are derived from housekeeping
proteins, it became interesting to understand why TEIPPs are not presented by processingintact cells. Convincing data shows that TEIPP peptides are actually produced within
processing intact cells, but somehow are not or not sufficiently presented by their surface
MHC-I molecules. Previous studies have shown that TAP-positive tumor cells can be
rendered sensitive to TEIPP-specific CTL after treatment with proteasome inhibitors
(e.g. lactacystin) or by deficiency of other APM components (e.g. tapasin) 1. Furthermore,
introduction of viral TAP inhibitors in DCs brings these internal peptides to the cell surface
for MHC-I presentation2. Taking the Trh4-derived TEIPP peptide as a model, we have
analyzed the expression of the Trh4 gene in several epithelial populations isolated from
wild-type and TAP1-ko mice (data presented in chapter 2). This analysis revealed the same
level of RNA transcripts between the normal and knockout populations, suggesting that the
proteins levels would also be comparable. The liberation of the Trh4 peptide is performed
by SPP which is active in TAP-positive as well as TAP-negative targets. Therefore, the
generation of the Trh4 peptide is comparable between normal and TAP-deficient cells.
But while APC from TAP-ko mice were recognized by Trh4-specific CTL, the wild-typederived APC were not. This shows that there is difference in the MHC-I presentation of
this TEIPP but not in expression. Additionally, enhancement of Trh4 expression in TAPpositive RMA cells by gene transfer, using vectors with a strong heterologous promoter,
resulted in presentation of the Trh4-derived peptide. Therefore, there is strong evidence
for the production of Trh4-derived peptide and other TEIPP in normal cells and that the
failure to present on normal cells has to be explained by failure to load in MHC-I.
What does prevent TEIPP from being presented in normal cells? One hypothesis is that
binding affinity of TEIPP peptides might be lower than other peptides and therefore
Discussion
141
loose competition with high affinity peptides that arrive in the ER via the TAP-transport.
However, we found that the Trh4-derived TEIPP peptide has a good binding affinity and
is capable of forming stable Db/Trh4 complexes, showing that it is an optimal MHC-I
ligand (data presented in chapter 2). Nevertheless, it is still possible that other TEIPPs are
indeed poor binders and that explains their absence from normal cells, but so far we did
not find evidence for that. To understand the absence of the Trh4-peptide from normal
cells we decided to look for the limiting step of Trh4 peptide presentation. We found that
increasing the availability of MHC-I molecules in TAP-deficient cells did not augment the
presentation of Trh4, indicating that TAP-deficiency per se creates the opportunity for all
available Trh4 peptides to be loaded, irrespective of MHC-I levels. This suggested that the
peptide quantity available for loading into MHC-I was the limiting factor. Indeed, increasing
the quantity of the Trh4 peptide in TAP-deficient cells increased the presentation of Trh4
and in normal cells resulted in the presentation of Trh4 whereas it does not normally
occur. Interestingly, increasing recognition by the TEIPP-specific CTL clone correlated
with fold of overexpression in target cells showing that the Db/Trh4 complexes increased
proportionally, implying that TAP function constitute a strong barrier for TEIPP. A vast
excess of TAP-pumped competing peptides is present in the ER and prevents TEIPP loading
to MHC-I. These findings argue that TEIPP peptides, such as Trh4, are underrepresented
in the ER and under normal conditions fail to find peptide-receptive MHC-I molecules in
the ER (Figure 1). We have confirmed that the C-terminal Trh4 peptide is liberated in the
ER, by the activity of SPP, as described in chapter 3. The study of a human TEIPP antigen
corroborates these findings. This antigenic peptide is encoded by the human CALCA gene
and derives from the signal sequence of preprocalcitonin (ppCT) protein. This peptide
is liberated in the ER lumen by sequential cleavage with SP and SPP, independently from
proteasomes and TAP3. The presentation of the ppCT peptide to specific-CTL was found
in human lung and medullary thyroid carcinomas (MTCs) that had very low expression
of TAP. Presentation of the ppCT peptide occurred also in normal non-transformed
cells, such as DCs, after knockdown of TAP. Overexpression of the CALCA gene in DCs
and TAP-positive tumor cells resulted in recognition by the specific CTL clone4. Thus,
these facts support the model of peptide competition in the ER as a factor that prevents
presentation of TEIPP antigens in normal cells.
Mechanism of peptide competition
We can speculate how such a peptide competition mechanism actually works. TAP is
one of the components of the PLC and links peptide transport into the ER with peptide
loading. These two processes physically linked by TAP enhance the efficiency of peptide
presentation. TAP-transported peptides readily get in contact with trimming enzymes
and chaperones needed for optimal peptide binding to MHC-I that are also part of the
PLC. The MHC-I molecules within the PLC have direct contact with the peptides that
come from the cytosol, a big pool that contains peptides with optimal binding properties.
Therefore these are favored over peptides situated in the ER lumen separate from the
142
7
Figure 1. Schematic illustration to explain why Trh4 peptides fail to be presented by processingproficient cells. We hypothesize that Trh4 peptides are underrepresented in the ER and lose competition
with the vast excess of TAP-pumped peptides under normal conditions. Under TAP-deficiency, those
peptides are not transported to the ER and Trh4 peptides get the chance to be presented instead. Binding
affinity and stability of the Trh4 epitope with its MHC-I molecule seem sufficient for presentation.
Adapted from: Seidel U.J. et al. A novel category of antigens enabling CTL immunity to tumor escape
variants: Cinderella antigens. Cancer Immunology, Immunotherapy 2012, 61(1): 119-125.
transport complex. As discussed above, peptides such as the TEIPP Trh4-derived peptide
are liberated in the ER in normal cells but are not loaded into MHC-I. The quantity of
these peptides seems to be too low to overcome the competition of the TAP-transported
peptides. In cases where the PLC does not have a functional TAP there is no direct
link between the entering peptides and the specialized loading complex. However, the
dynamical process of MHC-I presentation continues even without TAP. The available
peptides, probably situated in the vicinity of the PLC have the opportunity to be loaded
into MHC-I. Additionally, peptides liberated in the subsequent compartments of the
Discussion
143
secretory route might get in contact with peptide-receptive MHC-I molecules on their
way to the cell surface, which might exchange their ligands. Indeed, a higher number of
intracellular peptide-receptive MHC-I molecules is found in TAP-deficient cells 5.
Alternative ways to be loaded into MHC-I: find peptide-receptive MHC-I
in cells
Peptides located in the N- and C- terminus
The mechanism of peptide competition discussed here is a model that explains the
immunogenicity of TEIPP peptides, such as the prototype example Trh4-derived
peptide. This mechanism might be the base of the immunogenicity of other TEIPP that
are liberated in the ER similarly to the Trh4 and ppCT peptides. These are examples of
peptides situated in the N- and C-terminal regions of the proteins and can therefore be
easily liberated by mechanisms such as the processing of leaders and the action of SPP
on C-terminal regions of ER-resident proteins. The ER is equipped with sophisticated
machinery for optimizing ligand length and quality, and for facilitating peptide loading
onto nascent MHC-I molecules. This includes aminopeptidases as well as folding
chaperones and editing molecules in the peptide loading complex. In the absence of TAP
transporters the loading of peptides into MHC-I occurs without a fully functional peptide
loading complex. The details of this mechanism are not yet understood. It has been
shown that the loading of peptides in MHC-I can occur with the minimal components of
tapasin-ERp57-MHC-I complexes6. TEIPP peptides generated inside the ER may have
readily access to peptide-receptive MHC-I molecules associated with tapasin-ERp57 or
might be actively chaperoned towards these open grooves. Chaperones with high peptide
binding capacity in the ER are heat shock proteins (e.g. Hsp96) and PDI, an isomerase
that efficiently binds free peptide7. It is possible that TAP-independent peptides are
captured by these molecules and chaperoned to MHC-I.
Peptides located within proteins
A significant fraction of TAP-independent peptides identified thus far are derived from
internal regions of cytosolic proteins8, 9. These peptides, most likely, need the degradation of
the entire protein for being liberated in more complex degradation steps than the liberation
of N- or C-terminal ends. This normally occurs by the action of the proteasome, alone or in
combination with other proteases. We studied a TEIPP peptide antigen that was processed by
the proteasome and presented independently from TAP (chapter 3 of this thesis). In contrast
to Trh4, it is likely that in this case the peptide is generated outside the ER, via one or more
proteolytic steps. Novel transport mechanisms from the cytosol into vesicular compartments
that enable loading into MHC-class I molecules have been proposed. Apparently such
cytosol-derived peptides can have access to MHC-I loading compartments, such as the ER,
without the action of TAP. Several epitopes from the LMP2 protein of the Epstein-Barr virus
(EBV) are produced by the proteasome and presented without TAP10. It was suggested that
they cross the ER membrane by diffusion due to their hydrophobicity, because all of these
were highly hydrophobic. Maybe there are TEIPP peptides with such characteristics of being
144
generated in the cytosol by the action of the proteasome and then traverse membranes due to
their hydrophobic characteristics and do not need TAP.
Peptides generated in post-ER compartments possibly might travel back to the ER
for loading in MHC-I molecules. Peptides generated along the secretory route, including
Golgi, may use retrograde vesicular transport to get from distant organelles to the ER 11.
Secondly, MHC-I molecules with sub-optimally bound peptides may exchange them at
a second quality control check point in the Golgi12-14. Peptidic ligands generated in these
secretory compartments may bind to MHC class I molecules in these compartments
when the exchange occurs. It is likely that TAP-deficiency promotes the availability of
sub-optimally loaded MHC-I molecules and this increases the efficiency of presentation
of vesicular and secretory MHC-I ligands.
Autophagy as a way to load MHC-I?
Autophagy could provide a pathway to facilitate the loading of potential MHC-I ligands
in endovacuolar compartments. During autophagy, large portions of cytoplasmic content
are encapsulated in vesicles, the autophagosomes, and then fused with other vesicles
such as lysosomes. The contents of cellular organelles are degraded in these vacuolar
compartments in this conserved cellular process of self-eating. Antigens could be liberated
by proteases in a lytic vacuole after autophagy and encounter MHC-I complexes that
were recycled from the cell surface. It is known that MHC-I molecules constantly recycle
from the cell surface to early endosomes as part of the normal MHC-I trafficking. Upon
entering gradually acidic environments they might exchange their ligands. This vacuolar
pathway of antigen presentation might be of particular importance in TAP-deficient
cells or cells infected with viruses that can interfere with TAP or MHC-I processing and
transport. In a recent study it has been shown that autophagy promotes the processing
of viral antigens for TAP-independent MHC-I presentation. The MHC-I presentation of
an endogenous human cytomegalovirus (HCMV) latency associated protein, pUL138,
was followed the vacuolar pathway dependent on autophagy 15. Peptide loading occurred
within autophagic vesicles associated with lysosomes (autophagolysosomes). Curiously,
this pathway generated the same peptide epitope as that generated from conventional
processing. Thus, it is likely that this autophagy mediated pathway supplements the
conventional pathway and contributes to viral clearance.
The autophagosomal membrane has been proposed to originate from the ER 16, 17. If
the loading occurs in these compartments, the advantage created by TAP could be the
higher number of peptide-receptive MHC-I heavy chains that could enter other loading
compartments such as the autophagosome vesicles. However, more research is needed to
get further insight on the role of autophagy for MHC-I antigen presentation.
Conclusion
In summary, there are many new interesting pathways ready to be elucidated. And many
TEIPP antigens generated and presented via these pathways occur and stimulate the
immune system. Indeed the immunogenicity of TEIPP is an important topic to elucidate
Discussion
145
since it can be exploited in the combat of many diseases. The study of alternative pathways
could be exploited to develop new therapies to combat viral infections and cancer where
TAP is frequently inhibited. As we have shown, following TAP-inhibition there is a
generalized decrease in MHC-I presentation and the emergence of an alternative peptide
repertoire18. TEIPP peptides are among these TAP-independent peptides and might
constitute an important line of host defense for exploitation in therapeutic strategies.
The applicability of TEIPP antigens for immunotherapy of cancer has been tested in
preliminary studies using mice. The 9-mer TEIPP peptide Trh4 was used as a CD8+ T-cell
vaccine, combined with a T helper peptide to also stimulate CD4+ T-cells. This vaccination
scheme prevented the outgrowth of a TAP-deficient lymphoma1. These results showed that
TEIPP antigens can be used in combination with other antigens to be used as a vaccine that
can tackle tumor immune escaped variants. In the tumor vaccination field, several immune
intervention strategies have been designed and tested to stimulate the power and specificity
of the immune system to treat cancer. However, until recently, most of the immune-therapies
developed have provided poor tumor regression in humans. Some of these were based on
vaccination strategies to induce CD8+ T-cell responses against tumor-associated antigens
using single CTL epitopes comprising the exact HLA-I binding peptide. Peptide epitopes
used in clinical trials to treat metastatic cancer included melanoma-differentiation antigens
such as MART-1, gp100, tyrosinase or TRP-2 and cancer-testes antigens such as NY-ESO-1,
MAGE-12 or Her2/neu24. The low efficacy of vaccines using minimal CTL epitopes may be
caused by inefficient T-cell activation that precludes the ability of the T-cells to infiltrate the
solid tumors and become activated. In preclinical mouse studies it was shown that synthetic
peptides that were much longer and could reach up to 27 amino acids in length were more
effective25-29. These longer peptides may be retained in the draining lymph nodes where they
could be processed and presented by professional APCs to prime CD8+ T-cells. Additionally,
some of the synthetic peptides also contained CD4+ T-cell helper epitopes. These two
factors, processing and presentation of peptides by APCs and the induction of CD4+ T-cell
help are important factors that enhanced efficacy of these studies in mice but lacking in most
146
of the clinical trials in humans. This realization provided the development of even more
complex peptide formulations, for instance the synthetic long peptides (SLP). These contain
overlapping long peptides that span the whole protein and contain multiple CD8+ and CD4+
T-cell epitopes. A clinical trial was conducted in twenty women carrying pre-malignant vulvar
lesions associated with HPV16 infection using synthetic long peptides (SLP)30. This vaccine
formulation contained multiple MHC-I and MHC-II overlapping peptides (27-34 aminoacids) spanning the sequence of HPV16 E6 and E7 proteins. Five women had complete
regression of the lesion and this was associated with the induction of interferon--associated
CD4+ and CD8+ T-cell responses. Therefore, it might be worthwhile to test if we can formulate
long-TEIPP antigens to use as vaccines. One advantage of TEIPP antigens relative to tumor
associated antigens, such as differentiation antigens, it that they behave like foreign antigens
because they are not presented in the thymus in normal conditions. This implies that the
T-cells are not affected by tolerance which might contribute to the development of a robust
anti-tumor response and enhance the efficacy of tumor-specific immunotherapy.
Tumor-cell based vaccines
Description
Viral antigens
Point mutations
Examples
EBNA-1, E6, E7
MUM-1, CDK-4, p53,
Caspase-8
Differentiation antigens Expressed in tissue lineage
Tyrosinase, GP100, Mart-1
Cancer testis antigens Largely expressed during development and cancers MAGE, NYO-ESO-1
Cryptic epitopes
Associated with aberrant transcription and translation RU2, GnT-V, HPX42B
TEIPP
Associated with antigen processing defects
CALCA
Adapted from: Lampen M.H. et al, Current Opinion in Immunology 2011, 23(2):293-8
Discussion
147
TEIPP-specific CTL restricted by the non-classical MHC-I molecule Qa-1 are capable of
targeting TAP-deficient tumor cells such as other TEIPP-specific CTL restricted by classical
MHC-I molecules1, 2, 9. The particular characteristics of Qa-1 turn this molecule into an
attractive target for TEIPP studies. The Qa-1 binding groove shares a conserved structure
with the human homologue HLA-E47. Therefore the peptides presented by these two MHC-I
molecules would be very similar, implying that the discovery of Qa-1 presented peptides could
directly provide TEIPP epitopes to be used in human experiments. In fact, it was observed
that human cells transfected with Qa-1 were recognized by our mouse Qa-1-TEIPP CTL
clones (unpublished data) We found a surprisingly diverse peptide repertoire of more than
100 different peptides in Qa-1 on TAP-deficient cells, described on chapter 5. The peptides
originated from housekeeping proteins within the cell and some of them constituted neoantigens according to the TEIPP concept with the ability to give rise to CD8+ T cell responses.
Therefore a high number of antigens can be targeted in tumors by Qa-1-specifc CTL.
Qa-1 and HLA-E exhibit low polymorphism, consequently the presented peptides
are widely distributed within the population47-49. The independency of MHC typing on
the application of treatment strategies based on Qa-1/HLA-E-binding peptides contrasts
with most of the other MHC-I molecules. Interestingly, Qa-1-TEIPP CTL clones
recognized TAP-deficient tumors from different mouse MHC haplotype backgrounds
by virtue of the conserved nature of Qa-1. Therefore, Qa-1/HLA-E-presented TEIPP
antigens fulfill some requirements of ideal tumor antigens. As such, the potential use of
TEIPP peptides as vaccines is boosted.
Discussion
149
Moreover, Qa-1 shows a wide tissue distribution which allows the targeting
of different histological tumor types. One Qa-1-TEIPP CTL clone was capable of
recognizing colon tumors, melanoma and lymphoma (described on chapter 5). Qa-1
shows a relative maintenance of surface levels on TAP-negative conditions even in cells
that lost expression of classical MHC-I heavy chains47, 48, 50. This implies that tumors with
processing defects as well as tumors with MHC-I allele loss can be targeted by these
Qa-1-TEIPP CTL. In ovarian and cervical cancers, HLA-E expression was found in
equal or even higher levels than normal tissues. This was strongly associated with APM
deficiencies such as TAP-deficiency, HLA-I and HLA-II loss 51.
Conclusion
For more than 60 years it has been shown that the hosts immune system is capable of fighting
against cancer. One challenge of immune-therapy against cancer is the start of a robust and
specific CD8+ and CD4+ T-cell response that is capable of fighting against an established
disease. Tumors normally harbor immune suppressive environments and it might be useful
to relief the local suppressive milieu in combination with T-cell vaccination strategies. For
instance, tumors can be infiltrated with regulatory T-cells (Tregs) and myeloid-derived
suppressor cells (MDSCs)52-55. Therapeutic strategies that aim to break the tolerogenic
environment and induce inflammation might greatly enhance attraction of T-cells to the tumor
area and promote T-cell infiltration56. Therefore, the use of chemotherapy or radiotherapy
or other means that promotes the release of proinflamatory signals in combination with
immune-therapy might beneficial and are now focus of research57. For instance, TLR ligands
might increase the magnitude of T-cell response by strongly activate DCs and polarize it into
an inflammatory type of immune response (T-helper 1). Covalent linking of TLR ligands to
peptide vaccines is one strategy currently under study. Since the vaccine might also activate
pre-existing antigen specific T-regs it may be useful to deplete T-regs prior to the vaccination,
for instance by administering cyclophosphamide to the patients. These and other techniques
might help to improve the T-cell responses stimulated by therapeutic vaccines. TEIPP
antigens can be exploited to tackle T cell defense against processing deficient tumors and thus
be complementary to conventional tumor antigen vaccination.
| | REFERENCES
1.
2.
3.
150
Van Hall, T. et al. Selective cytotoxic Tlymphocyte targeting of tumor immune escape
variants. Nat. Medicine 12, 417-424 (2006).
Chambers, B. et al. Induction of protective
CTL immunity against peptide transporter
TAP-deficient tumors through dendritic cell
vaccination. Cancer Res 67, 8450-5 (2007).
El Hage, F. et al. Preprocalcitonin signal peptide
generates a cytotoxic T lymphocyte-defined
tumor epitope processed by a proteasomeindependent pathway. Proc Natl Acad Sci U S A
105, 10119-24 (2008).
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
151
152